Drivers of Gendered Sectoral and Occupational Segregation in Developing Countries

ABSTRACT

Occupational and sectoral segregation by gender is remarkably persistent and a major contributor to gender wage gaps. We investigate the determinants of aggregate occupational and sectoral segregation by gender in developing countries using a unique, household-survey-based aggregate cross-country database including sixty-nine countries between 1980 and 2011. Using two aggregate measures of segregation the study shows that occupational and sectoral segregation has increased over time in many countries. It finds that income levels have no impact on occupational or sectoral segregation; trade openness has little impact on sectoral segregation but increases occupational segregation. Rising female labor force participation is associated with falling sectoral but increasing occupational segregation; rising education levels tend to increase rather than decrease segregation. Income inequality is associated with rising segregation. While the overall effects of structural change are small and mostly insignificant, a high share of commerce and services is associated with lower overall segregation.

KEYWORDS:

• Occupational segregation
• sectoral segregation
• gender
• developing countries

JEL Codes:

• J16
• J24
• J71

ACKNOWLEDGMENTS

We would like to thank Diane Elson as well as participants at workshops at the New School in New York, the University of Göttingen, the IMF, and the University of Stellenbosch, as well as three anonymous referees for helpful comments on earlier versions of this paper. We would like to thank Claudio Montenegro from the World Bank for providing access to the I2D2 database.

SUPPLEMENTAL DATA

Supplemental data for this article can be accessed at https://doi.org/10.1080/13545701.2019.1649708.

Notes

1 It also includes transition countries in Eastern Europe and the former Soviet Union, except those that are members of the Organisation for Economic Co-operation and Development (OECD).

2 See a related theory by George A. Akerlof (1997) on social distance that could also be applied to occupational or sectoral segregation.

3 See also related theories on “queuing” by Barbara Reskin and Patricia A. Roos (1990), and “relative attractiveness” of occupations by Myra H. Strober and Lisa M. Catanzarite (1994).

4 We would like to thank Claudio Montenegro from the World Bank for generating and providing the aggregate data for our analysis.

5 Tables 11 and 12 show how many observations we have per country.

6 We do not consider rural areas where agriculture is usually the dominant sector and occupational distinctions are small and not easily comparable across countries.

7 Jacobs and Lim (1992) refer to this as the “parallel lines” hypothesis and find it holds for nine out of ten countries for which they test this hypothesis.

8 Also a combined sector of the following five service categories: public utilities, transport and communications, financial and business-oriented services, family and community-oriented services, and other services

9 We dropped the agriculture employment share as the omitted reference category.

10 Using the Gini for disposable income leads to very similar results (available on request).

11 Remember that the max for the IP is 0.50 and values will be exactly half of the ID when there is equal male and female labor force participation.

12 We tested this by calculating the means for the sectoral ID and IP in a restricted sample of non-missing values of their respective occupational index values and the resulting means converged to 0.24 and 0.12, respectively.

13 A Hausman test was performed to confirm the use of a fixed effects rather than a random effects specification.

14 In addition, we cannot capture if fixed unobserved factors have a time-varying impact.

15 Note that this trend does not refer to a uniform year range or number of years, given the unbalanced nature of our data set. Any country for which there was more than one year of index calculation available was included here. Not all countries exhibited clear trends over time, either, as in some there were fluctuations. For consistency, the first and last year values were compared regardless of any fluctuations in between.

16 For consistency we performed additional regressions restricting our sectoral ID and IP sample to non-zero values of the occupational ID and IP. This did lead to changes in magnitude and significance of some of our coefficients, but not to changes in signs and did not eradicate the differences in results between sectors and occupations, so we include the full sample regressions here.

17 The difference in results in the specification with and without time fixed effects (columns 3, 4 and 7, 8 in Table 4) is smaller but still notable.

18 As argued above, it is unlikely that biases caused by endogeneity would generate a positive correlation.

19 As expected and shown in the Online Appendix, the coefficient is larger for both regressions when IP is used, given its in-built sensitivity to female participation; but also there, greater female participation significantly lowers sectoral and increases occupational segregation. As discussed above, this result might partly be affected by reverse causality where rising sectoral segregation discourages female participation so that we should interpret this coefficient with caution

20 Since this effect is not significant for the SIP index, it appears to be driven by large sectors.

21 They argue that in these societies, the provision of subsistence and basic rights increases the scope for gender-specific ambitions and desires.

Additional information

Funding

Funding from the Hewlett Foundation, the British Department for International Development, and Canada’s IDRC as part of the GrOW (Growth and Economic Opportunities for Women) project is gratefully acknowledged.

Notes on contributors

Mary Borrowman

Mary Borrowman holds a PhD in economics from the New School of Social Research.

Stephan Klasen

Stephan Klasen is Professor of Development Economics at the University of Göttingen. He is also the director of the Ibero-America Institute for Economic Research and the Coordinator of the Courant Research Center “Poverty, Equity, and Growth in Developing and Transition Countries.” He holds a PhD from Harvard University and has since held positions at the World Bank, King’s College (Cambridge, UK), and the University of Munich. His research focuses mostly on issues of poverty, inequality, environment, and gender. He is a member of the UN Committee on Development Policy, and was President of the European Development Research Network.